Therapy transcripts where identity negotiations were evident were analysed to inductively generate two main themes with embedded subthemes to capture the identity negotiations evident as the participants engaged in CBT-AN/LEAP or CBT-AN only.

Theme 1:

Troubled Identities

Theme 1a:

Conflicted Identities

Theme 1b:

Identities as Othered

Theme 2:

Rebuilding Identities

Theme 2a:

Shifting Relationship with Oneself

Theme 2b:

Building Life and Identities Outside of the AN Identity

The inter-relationship between these themes and subthemes (Fig. 1) was recursive with the person returning to the AN identity, albeit less frequently, over time. See Additional File 1 for all participant extracts.

Fig. 1Theme 1: troubled identities

When therapy sessions focused on identity talk, the participants’ identities were troubled by the AN identity and what this meant to them as a person. At these points in therapy conversations, the AN identity was both conflicted (Theme 1a) and othering (Theme 1b) of the person with a lived experience.

Theme 1a: conflicted identities

The range of identity positions taken up by participants when talking about the AN experience over the course of treatment varied and had implications for the extent by which the AN identity was conflicted. This extended to a struggle in navigating a sense of identity both with and without the presence of AN.

Early therapy sessions

Within early therapy sessions, participants took up a variety of conflicted identity conclusions. For example, Participant Y concluded that they ‘…felt really weak… I need to stop being such a baby’ (S1), and Participant V that they felt ‘…really dirty’ (S1). These negative identity conclusions appeared emblematic of a broader struggle for participants to live with and accept themselves. Furthermore, participants negotiated to what extent their identity was interwoven with AN, including questioning whether AN was them, a part of them, or separate to them. For Participant Y, the troubling identities generated by AN led to a loss of agency to AN whereby they became ‘…dissociated from a rational person and you just have to act that out because you’re possessed by it’ (S1). This contributed to a sense of separation from the self and disintegration of Participant Y’s identity outside of the AN identity. The impact of this was a picture of a divided sense of self that was mapped along the fault lines of the AN experience.

Participants S and V resisted the idea that the ED was separate from their sense of selves (“separation externalisation”) [60] (p. 139). Instead, Participant S took up the position that ‘There are definitely two sides of me, the rational side of me and the ED. I really do not feel like there are two different voices or things like that’ (S1). This suggested that whilst Participant S experienced different parts of herself not occupied by the ED, there were also aspects of her identity that included AN that were positioned as “two sides of me”. For Participant V, AN was not positioned directly as separate to the self, however, they drew on externalised language when explaining how ‘it [AN] has control’ (S3). Alternatively, for Participant Q, her own research into AN led her to position herself outside the AN diagnostic category, despite also seeking treatment, i.e., ‘I research all that stuff… I just don’t see myself as someone who has it’ (S1).

The conflicted nature of AN was also evident at points when participants argued that AN was protective. Participant Z noted that the pain of compulsive exercise was experienced to her as a ‘good thing’ (S2), and Participant R described AN as ‘familiar’ and ‘secure’ (S3). Within the same stretch of text, an antithetical position was taken up where participants appeared to be arguing within themselves against a life dominated by AN, which was described by Participant Z as ‘running myself into the ground’ (S2) and ‘destructive’ (S9) by Participant R. At this point in therapy, these discussions were often dialectical in nature where participants appeared to attempt to resolve these seemingly conflicting positions.

Mid-therapy sessions

Although there appeared to be less AN-dominated negative self-concepts expressed during mid-therapy sessions, participants continued to negotiate identities that were shaped by societal values of the thin body (i.e., thin). Participants S and Y described a sense of failing to meet the cultural ideals of the thin and exercise shaped body as leading them to see themselves as ‘lazy’ (S6; Participant S) and ‘not even as good as any of these people’ (S9; Participant Y). Implicit in these extracts was a moral discourse whereby participants’ sense of self-worth was measured by their engagement in exercise and attainment of the thin body.

Likewise, participants appeared to enter more nuanced and complex discussions regarding their experience of AN as being both separate and intertwined with the self. Participants’ positioning on this question had implications for the extent to which they experienced personal agency over AN. This included discussions on whether they perceived AN to be a form of control, to be controllable, and/or in control of them. The nuanced nature of these identity negotiations was exemplified by Participant V who stated, ‘I don’t know if that’s all driven by the ED, or it could just be my personality’ (S12). Within this quote, Participant V first took up a position that separated the ED from the self, whereby this meant the ED took on a life of its own. They then took up a conflicting position that internalised the ED as part of her personality. Both positions had the effect of diminishing the participant’s sense of personal agency. This dialectic around the question of personal agency was also evident for Participant X, who first described AN as having total control over their life—‘It dictates everything’ (S10). Alternatively, later in the same session, Participant X reflected that regardless of this they ‘…don’t want to let it [the ED] go. I just cannot let it go’ (S10).

For Participants V and X at this point in therapy, the uncertainty of living with the self without AN appeared overwhelming and a dual relationship with AN emerged. Participant V highlighted the conflictual nature of the AN experience with the recognition that AN was causing her distress but also that she perceived that AN ‘…helps me to live with myself’ (S12). Likewise, Participant X expressed a fear that ‘pain and stuff’ (S10) might come to the forefront of her life in the absence of AN. Implicit in this was a function of AN to push back past painful experiences which made it challenging to ‘let go’. Moreover, these difficulties in having agency to let go of AN were implicit in Participant Z’s desire for the AN to ‘go away’ (S15). Participant Y highlighted further complexities in navigating ideas of self-agency or control over AN, evident in her concern that ‘As I try to control the [restricting] more, I wonder if I’m going to try and be more controlling of other parts of my life’ (S9). This speaks to a conflictual relationship with agency and control, in particular Participant Y’s concern that increased control over AN symptoms may lead to her wanting to control other aspects of her life more, perhaps in unhealthy ways.

Late therapy sessions

The struggle with conflicted identities related to the AN experience was ongoing for some participants in later therapy sessions. Participant X continued to draw on a moral discourse that meant her hopes and desires were confined to the attainment of the thin body—‘Everything you want is to be skinny’ (S20). Implicit in this quote is Participant X’s previously expressed struggle to live with themselves should they not be thin. Additionally, Participant R described AN as a protective layer to a ‘fake’ identity that lay underneath—‘… people will see through me… Like I’m a fake’ (S22). The vision of recovery for Participant R was one of subscribing to societal norms and ‘…trying to do everything I could to mimic what everyone else was doing to fit in. To not look like a fraud’ (S22). Without AN, Participant R imagined they would be left with a fraudulent identity that would eventually be exposed to others. This highlighted the difficulties experienced by this participant in letting go of AN that were intertwined with an underlying impoverished self-concept.

Participants’ accounts in later therapy sessions continued to highlight the complex ongoing identity negotiations in relation to AN, in particular the extent to which each experienced themselves as having personal agency over AN. For example, in positioning the ED as separate to herself, Participant W’s narrative that she had not ‘…been in control for a long time’ and that ‘The ED has been in control’ (S30) continued to be reinforced. On the other hand, Participant R’s account of her identity in relation to AN was internally conflicted between a part of her that ‘…knew that I had to put on weight and I had to change my behaviour’ and another part that felt making these changes was ‘…connected to losing control’ (S22). Participant S’ identity in relation to the AN experience shifted later in therapy, which was evidenced by her comment that it [AN] was no longer ‘…the same anorexia I was back then’ (S26). This suggests that while AN continued to have a role in her self-conceptualisation, this role had shifted and her personal agency over AN had increased, with a sense that Participant S was standing for herself at this point in the therapy—‘I definitely would do something to correct that [if ED symptoms returned] because I know I am there for me’ (S26). Whilst participants continued to grapple with a sense of personal agency in their relationship with AN, what was also evident was their increasing insight into their internal identity negotiations and considerations of how they might like their relationship with AN to be different.

Theme 1b: identities as othered

All but one of the participants spoke about a sense of feeling othered by the AN experience through (i) the sense of isolation from friends and family relationships who struggled to understand their experience, and (ii) being positioned as outside the boundary of normality.

Early therapy sessions

Early session transcripts indicated that participants’ experiences of AN were often associated with a sense of being othered and leading to isolated identities. AN was talked about as contributing to a sense of alienation and separation from others in their life, including their loves ones. Participant X felt as though their family members did not understand their experience of an ED, stating, ‘…he does not get my own issues… he does not get my point’ (S1). Implicit in this was the experience of an isolated identity generated by the ED that others could not access. A unique experience of an isolated identity was further described by Participant S who had a sister also struggling with AN. Comparisons with her sister’s perceived recovery contributed to strain in her family relationships where she did ‘…not want to prove myself to people who do not realise how difficult it is to maintain weight’ (S6). When Participant W reflected on how the AN had contributed to the impacts of her actions on others, the most available identity position reduced her sense of herself as a person to ‘… short tempered and impatient and just a screaming mother’ (S1). A further issue of blame was raised by Participant V who spoke of her family members defending their identities against blame for the AN, i.e., ‘They get really defensive. They think it’s all their fault’ (S1).

Early therapy sessions also explored the impact uptake of discourses of normality had in shaping participants’ identities as other than ‘normal’. Participant X argued that although ‘Every day I wanted to work out, I never wanted to be crazy’ (S1). This indicated that in taking up the AN identity, there were limited identity positions available outside of a pathologized and disorganised identity. On the other hand, Participant Z described a resistance to therapy that she experienced as ‘… trying to normalise me’ (S2), including an over focus on food and journaling her meals, stating, “It does not feel normal. Normal people do not write down food’. (S2). She instead argued that ‘food is not the issue’ and that she needed treatment to focus on the issues that were relevant to her. Participant V also positioned herself outside normality when noting her ‘best method’ for recovery was to ‘line myself up with an engagement with someone… And so that way I go to control myself to being normal’ (S8). Implicit in Participant V’s account was a fragmented identity outside of AN where normality was something to be controlled into by mirroring the eating behaviours of others not suffering from an ED. The inherent challenges of defining so called normality meant that this was construed as a vague and distant goal that could not be well described.

Mid-therapy sessions

Discussions of relational difficulties and how these contributed to isolated participant identities continued in mid-therapy transcripts. Participant X described AN as affecting ‘…your relationships and it puts a strain on things… [the ED] makes you have more arguments with people more often because they don’t understand you… only you understand yourself and it’s just easier to be alone’ (S10). For this participant, the strain caused by AN and profound feelings of being misunderstood led to a resignation towards ongoing solitary identities (i.e., ‘easier to be alone’). This sense of isolation contributed to an identity as othered as exemplified in her description: ‘…instantly feel[ing] bitter towards really thin people, because I wish I was them’ (S10). For Participant T, isolated identities were driven by not wanting others to ‘see me eating’ (S14), which limited her capacity to socialise and connect with others.

Participants continued to be positioned by discourses of normality in mid-therapy transcripts. Participant V described a process of ‘trying to adopt more of a normal person’s perspective in arguing with the disorder’ and that ‘…because there were other people around, it was probably more important that I looked or appeared to be normal’ (S12). These extracts show how through positioning herself outside the boundary of normality, becoming ‘normal’ was construed as required for the motivation or goal for recovery. Likewise, these statements positioned normality as something required to be socially accepted by or connected to others. Participant Y indicated that she felt regular exercise was fundamental to her self-worth and ability to meet ‘normal’ standards—‘If I’m not doing exercise then I’m not even as good as any of these people. I’m like in a different level of society’ (S9). Participant W’s account exemplified an internal dialogue of shifting positions between seeing herself as ‘normal’, ‘huge’, and then ‘not anorexic anymore’, i.e., ‘I feel normal. I feel huge really […] I’m 44kgs, that’s a normal person weight, not an anorexic weight. I’m not anorexic anymore […] I’m seeking treatment but I’m not anorexic’ (S14). The implications of this were a conceptualisation that a person’s weight could fall inside or outside the bounds of normalcy and the implications (perceived or real) for her legitimacy to continue to engage in ED treatment should she or her weight be ‘normal’.

Late therapy sessions

Ongoing fears of being judged or rejected by others were evident in therapeutic conversations in later sessions, contributing to a continued sense of isolated identities. Participant R expressed concerns about ‘worry[ing] or burden[ing]’ others, and ‘the chance that the new people won’t like me or will reject me’ (S22). Likewise, Participant Z disclosed that they were ‘…normally stressed about if my new friends are judging me…’ (S32). For Participant W, isolated identities generated by AN continued to be exacerbated by a sense that others ‘… do not understand what it [AN] is’ (S30). Overall, the extracts from end-of-therapy sessions highlighted the enduring sense of identity as othered that persisted and was exacerbated by struggles to feel understood and emotionally safe in relationships, fears of criticism, and the sense of oneself as not normal.

Theme 2: rebuilding identities

How participants depicted their shifts in their relationship with AN had implications for who they understood themselves to be and how they conceptualised recovery. These shifts included (a) shifting relationship with oneself and (b) the building of life and identities outside of the AN identity through finding previous identities lost in the AN experience.

Theme 2a: shifting relationship with oneself

Most of the participants spoke at some point across the three therapy sessions about what recovery meant to them individually and their vision of themselves as ‘recovered’. The nature of these identity negotiations fluctuated over treatment.

Early therapy sessions

Within early therapy sessions, participants were able to identify hopes and dreams for recovery. Rather than characterised by a reduction of symptoms or increased weight, recovery at this point in the therapy was spoken about holistically as an embodied shifting relationship with oneself. This included Participant Y who identified that her ‘…greater goal would be… just relaxing into the way I am a little more’ (S1). Participant R conceptualised recovery as allowing the self to ‘accept the practical and rational reasons for taking it easy sometimes’ and ‘allowing the body to recuperate’ (S4). This greater connectedness to the self was also envisioned as an improved responsiveness to the body. Likewise, Participant X discussed the importance of recovery being about having a ‘happy and healthy mind’, rather than a sole focus on weight or symptom reduction—‘There is a difference between a healthy mind and a healthy body. My body wants to be healthy and happy, but where is my healthy and happy mind?’ (S1). For Participant Q, ‘self-esteem [was] the main thing’ (S1) involved in recovery. Implicit in these extracts were participants’ abilities to use complex and nuanced descriptions of recovery that was fundamentally related to a shift in relationship with the self. Simultaneously, participants experience of this connection with the self was that it had been disrupted by the AN experience.

Mid- and late therapy sessions

In transitioning from early therapy to mid- and late therapy sessions, discussions conceptualising recovery appeared to become more troubling for participants. Where once able to clearly identify descriptions of recovery from AN, participants appeared to struggle to generate these preferred ways of being, particularly in the context of a medicalised version of recovery where the illness is assumed to be gone. Participants experienced difficulties imagining their life on the terms of this medical discourse, where the only available identity slots are either sick or recovered. For Participant Z, this was expressed in her statement, ‘I don’t know what recovered is like, I don’t know what it feels like’ (S15). Likewise, Participant V considered that ‘Maybe if I got better at this… maybe the stronger I believe this, the more likely it’s going to “beat” the ED itself’ (S21). These extracts demonstrated the emphasis participants placed on their hopes for eradicating AN from their life, and the sense of a distant and undefined sense of identity that came with these hopes.

Theme 2b: building life and identities outside of the AN identity

Whilst some participants struggled to conceptualise recovery and what this meant to them, there were stretches in the therapy sessions over time for the majority of participants where shifts in their relationship with AN appeared possible. This was through a shift in focus from eliminating the AN symptoms to building the sort of life and identities they hoped for.

Early sessions

Early in therapy sessions, participants found it difficult to engage with conversations around building life and identities outside of AN. Instead, therapeutic dialogue focused more on exploring the role, function, and impacts of AN on the person’s life. During their first therapy session, Participant X, however, was able to identify aspects of her life that she wanted to reclaim from the ED and that spoke to her values and what mattered to her—‘… I used to play a lot of music and write things… I want to go back to only keeping myself busy with things that make me happy and not the ED […] I really like it that I can slow down, sit with my kids, you know this has really started to come back’ (S1).

Mid- and late therapy sessions

Therapeutic conversations throughout mid- and late therapy sessions saw participants more frequently explore how shifts in their relationship with AN would impact their identities and ability to rebuild their lives. For Participant S, rebuilding life was an acknowledgement that ‘I can add another chapter in my life’ (S13) which she described as ‘very exciting’. The language used by Participant S in this quote of ‘I can’ highlighted an increased sense of agency and ownership over the direction of her life. Likewise, Participant W described recovery as an increased focus on the rest of her life outside the disorder and ‘letting go’ of AN—‘I just want to start to be able to not think about it and want to get on with my life without giving it focus. I want to start letting go’ (S14). Central to Participant W’s preferred sense of identity were her values and re-establishing a connection with herself and others with an externalisation of AN through the use of ‘it’. This was also evidenced by a journal extract that Participant W read out to her therapist which depicted a connection with a valued sense of self—‘Calm and happy, nice, patient, kind, logical, healed, transmit health, compassionate, beautifully strong and fit, relationships, healthy, balanced, happy kids, help others, freedom, picnics, going out for dinners, loving self, socially free and liberated’ (S14). This journal had scope to provide Participant W with a map to navigate her identity as she migrated from a life and identity dominated by AN. Nearing the conclusion of treatment, Participant Z indicated that she felt she was ‘Working together for the goals… and crossing them and building a future’ and that ‘it feels so nice to be here’ (S32). This conceptualisation of recovery was linked to a sense of rebuilding a life outside of AN, characterised by a greater knowledge of, working towards, and achieving her personal goals. Likewise, Participant S described identity shifts in terms of increased independence and confidence, in addition to taking back control over her choices—‘… a lot more independence […] You have more control over how you manage meals and everything’ [26].